3.2 Case characteristics
The case organization operates in the modular construction business and delivers inte-grated building solutions mainly for the procurement of the public sector. Common prod-ucts are schools, daycare centers, and offices. The case organization is a Finnish sub-sidiary of a Northern European corporation.
Compared to more traditional, fixed construction business, modular buildings are con-structed using prefabricated, relocatable modular elements. After the procurement deci-sion, modules are transported and assembled to construction sites for the contract pe-riod. After the customer need ceases to exist, modules are disassembled and re-used in another building. The average area of a building module is approximately 30 square me-ters, which equals roughly to a large shipping container.
Building delivery projects vary considerably by the size and cost of the building and, therefore, by their significance to the case company. The delivery of multiple building projects happens simultaneously, and the case organization needs to prioritize re-sources constantly within the project portfolio. To retain the effectiveness and to be able to forecast and report the performance of projects, project portfolio, and ongoing period, the case organization is required to estimate the resource expenditure of its operations.
The case organization is structured as a matrix organization. The employees are split between management, sales department, project department, and fleet department.
Each department operates with each product and each project, which highlights the im-portance of collaboration throughout the organization. An outlined organizational chart based on the researcher’s experience in the organization is presented in Figure 2. On top of the business functions that are presented in the figure, the case organization is supported by group functions, such as HR and IT.
Figure 2. Organizational chart of the case company.
It should be noted that the researcher has been working in the case organization in dif-ferent roles for multiple years and therefore is an active practitioner in the case organi-zation. Whereas this may enrich the data gathering, this may affect the objectivity of this study.
3.3 Data gathering
The qualitative data for this study was collected by interviewing the employees of the case organization. The objective of the interviews was to gather comprehensive insight into the cost estimation process and estimation methods, surrounding activities, issues, future development plans, and improvement ideas. For this objective, interviewees were selected using purposive sampling to interview the most impactful actors related to the cost estimation activities. Purposive sampling enables the use of the researcher’s judg-ment to select cases that will best enable answering the research questions (Saunders et al. 2009, p. 237). The purpose of the sampling was to interview employees that would either
• Carry out cost estimation activities of modular building projects
• Manage certain actions in building delivery project as process owners
• Be involved in the development of cost estimation processes, methods, or tools.
Interviews were carried out as semi-structured, interviewing employees from different organizational roles and operational functions. Interview structure and questions were
generated mainly on the researcher’s reflections on what would be required to answer the research question. Some of the interview questions were adapted from Rush & Roy (2001). The interview structure on an aggregated level is presented below. The entirety of the interview structure is included in appendix A.
• Background of the interviewee
• Objectives of cost estimation
• Current cost estimation process
• Cost estimation tools and techniques
• Planned development in estimation and surrounding activities
• Risk and uncertainty management
• Issues regarding cost estimating activities
• Assessment of the current performance of cost estimating
• Suggestions for optimal state of the cost estimation process
Ten different employees were interviewed on 11 occasions between April 2020 and June 2020. In each of the interviews, the researcher was the interviewer, and all interviews were only between the interviewee and the researcher. Six interviews were held via video call due to the company’s meeting restrictions caused by the global pandemic situation.
All interviews were held in Finnish, which was the native language of all the interviewees.
Information on interviews is presented in Table 4. All interviews were recorded, and all interviews were completely transcribed.
Table 4. Interviewee information and interview durations.
Code Position
Interview length [min]
Interviewee's experience in the case company
H1 Development manager Upper management 25 4 years
H2 Head of the project department [#1] Upper management 65 6 years
H3 Sales representative Worker 45 5 years
H4 Head of the project department [#1] Upper management 71 6 years
H5 Managing Director Senior management 57 3 years
H6 Head of the sales department Upper management 79 3 years
H7 Head of the project department [#2] Upper management 36 <1 year
H8 Site manager Worker 36 25 years
H9 Project manager [#1] Middle management 55 2 years
H10 Project manager [#2] Middle management 32 2 years
H11 Business controller Upper management 66 4 years
Average: 52 min Total: 567 min
The interviewees came from different organizational roles, mainly from business unit management, sales department, or project delivery department. Also, the working expe-rience in the case organization ranged from under one year to over 25 years of experi-ence. These characteristics challenged the use of the interview structure as is because the interviewees had vast differences in their expertise areas.
For empirical research, another data gathering method was investigated but was aban-doned. Cost estimates and project follow-up from past projects could have been ana-lyzed to investigate cost estimation methods, documentation, and estimation accuracy.
Resulting from internal development that has targeted especially the tendering practices, it was analyzed that comparison of cost estimates between these projects would have been a considerable challenge. Also, due to the lack of definition in some of the estima-tion activities, the level of documentaestima-tion would not have enabled systematic analysis without additional interviews in the case organization.
3.4 Data analysis
Each interview was completely transcribed with the assistance of a digital transcription audio player (Express Scribe). Transcriptions were enhanced with verbatim elements.
The interview data was treated with multiple different coding approaches. After the first read-through, a chronological outline for the cost estimation process and its aggregated phases were generated. First level coding was conducted based on this outlined pro-cess, and meaningful data for cost estimation activities were categorized in these phases. The reasoning behind first-level coding was to compress data into smaller sec-tions that would be faster to analyze.
The interviewees were eager to discuss issues that they are experiencing when perform-ing cost estimation activities in daily operations. These issues were related to all parts of the assembly project, and they were caused by several reasons. To utilize these issues more structurally, pattern coding was conducted, utilizing a method adapted from Gioia et al. (2012). The reasoning behind this coding was to apply analytical discipline to the data and to make it more easily interpretable. Also, since one objective for this study is to find out issues in current practices, the data model presented acts as an intermediate product to accomplish this objective.
The data model in the Gioia method has 3 levels, first-order concepts, second-order themes, and aggregate dimensions. First-order concepts consist in this case of the ex-perienced issues in building projects related to cost estimation activities and surrounding
processes. These issues were marked in the data, which provided 44 first-order con-cepts. The second-order analysis aims to group related findings into categories that are given a name, an explanation, which limits the categories to 12. Finally, these themes are grouped in a handful of workable aggregate dimensions that enable analyzing and addressing the issues on an aggregated level. The data model also acts as a visual aid in how the data were categorized. The data model is presented in chapter 4.3.